Journal of Colloid and Interface Science 311 (2007) 375381

www.elsevier.com/locate/jcis

Adsorption of phosphate and nitrate anions on ammonium-functionalized

MCM-48: Effects of experimental conditions
Rabih Saad, Khaled Belkacemi, Safia Hamoudi
Department of Soil Sciences and Agri-Food Engineering, Universit Laval, Laval, QC G1K 7P4, Canada
Received 3 January 2007; accepted 14 March 2007
Available online 23 April 2007

Abstract
In this study, ammonium-functionalized MCM-48 (Mobil Composite Material No. 48) was used as an adsorbent to remove nitrate (NO
3 ) and
)
anions
from
aqueous
solutions.
The
effects
of
operating
conditions
such
as
temperature,
adsorbent
loading,
initial
monobasic phosphate (H2 PO
4
anion concentration, pH, and the presence of competitive ions on the adsorption performances were examined. Results showed that adsorption
capacity decreased with increasing temperature. The adsorption capacity increased with adsorbent loading and initial anion concentration. The
removal of nitrate was maximum at pH < 8, while phosphate removal was maximized at pH 5. The adsorption was almost unaffected by the
presence of competitive ions in the case of phosphate anions. However, their presence adversely affected nitrate adsorption. Desorption of both
anions was rapidly achieved within 10 min using NaOH at 0.01 M. Regeneration tests showed that the adsorbent retained its capacity after
5 adsorptiondesorption cycles.
2007 Elsevier Inc. All rights reserved.
Keywords: Adsorption; Nitrate; Phosphate; Ammonium-functionalized MCM-48; Desorption; Regeneration

1. Introduction
Water pollution due to the excessive presence of nutrients,
i.e., nitrogen and phosphorus species, is a serious environmental problem worldwide, as both are implicated in eutrophication
of receiving surface waters [1]. Agricultural overapplication of
natural and synthetic fertilizers, aquaculture, municipal wastewaters, detergent manufacturing as well as mineral processing industries are the main sources of nutrient release into the
aquatic environment [2]. Eutrophication comprises the abundance of aquatic plants, growth of algae, and depletion of dissolved oxygen [3]. Furthermore, the presence of nitrate ions in
drinking water is a potential public health hazard and can cause
infant methemoglobinemia, blue baby syndrome [4,5].
Several physicochemical and biological processes have been
investigated for the removal of dissolved nutrients in water and
wastewaters. Among them, adsorption methods are promising
* Corresponding author. Fax: +1 418 656 3723.

E-mail address: safia.hamoudi@sga.ulaval.ca (S. Hamoudi).

0021-9797/$ see front matter 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2007.03.025

since they allow simple and economical operation resulting in

less sludge production and fewer disposal problems. Therefore,
several materials such as fly ash and cement [6], surfactantmodified zeolites [7], alunite [8], polymeric ion exchangers [9],
and agricultural residues [10] have been investigated as adsorbents for the removal of phosphate and/or nitrate anions.
However, these materials exhibited insufficient adsorption and
regeneration capacities as well as poor selectivity and limited
surface area. Therefore, it is imperative to develop novel adsorbents with high surface areas and open pore structures for
enhanced rate adsorption and increased adsorption capacity.
During the past 2 decades, the so-called mesostructured materials received a great deal of attention from the scientific
community. Several strategies were proposed to assign specific
properties to these materials for their adequate use in a variety of applications such as catalysis, adsorption, and chemical
sensing [11]. Among these strategies, surface functionalization via tethering of organic functional groups attracted much
attention, as it allowed a number of functionalities to be introduced onto the internal pore surfaces through cocondensation or postsynthesis grafting. The grafting method consists

376

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

in reacting a suitable organosilane with a previously synthesized mesoporous material using an appropriate solvent under
reflux conditions. Thus, the surface silanol groups react with
the organosilane to form a layer of covalently coupled surface
functional groups. Cocondensation involves the hydrolytic condensation of an organosilane with conventional silica sources
such as tetraethyl orthosilicate. Specifically, mesoporous silicas
functionalized with amine functional groups have been applied
as adsorbents for heavy metal cations, such as cobalt, copper,
and zinc cations [12], as well as for chromate and arsenate
oxyanions [13].
The present work is intended to study the adsorption of nitrate and monobasic phosphate anions from synthetic aqueous
solutions using a surface-modified mesoporous silica, MCM48 (Mobil Composite Material No. 48), as a novel adsorbent.
The effects of adsorbent loading, initial anion concentration,
temperature, pH, and the presence of competitive ions on the
adsorption performance were investigated. Desorption kinetics
and regeneration were also examined to explore the practical
usefulness of the adsorbent and its possible use in a continuous
sorption process.
2. Experimental methods
2.1. Chemicals
Tetraethyl orthosilicate (TEOS; purity >98%), 3-aminopropyltriethoxysilane (APTES; purity >99%), cetyltrimethylammonium bromide (CTAB; purity >99%), ammonium hydroxide (NH4 OH), and sodium nitrate (NaNO3 ) were purchased
from Aldrich. Sodium orthophosphate (NaH2 PO4 ) was obtained from BDH. All chemicals were used as received without
further purification.
2.2. Synthesis of MCM-48NH+
3
MCM-48 was prepared according to the synthesis procedure
described by Schumacher et al. [14]. The gel composition was
TEOS:0.4CTAB:1.5NH3 :54ethanol:174H2 O. In a representative synthesis, CTAB (2.4 g) was dissolved in 50 ml deionized
water, then 50 ml of ethanol was added. The pH was raised to
ca. 12 by adding NH4 OH, followed by the addition of the appropriate amount of TEOS. The resulting solution was stirred
for 2 h at ambient temperature. The obtained white precipitate
was filtered, washed thoroughly, and dried at 323 K.
The functionalization of MCM-48 leading to amino-functionalized (MCM-48NH2 ) was conducted via the postsynthesis
grafting method according to Yokoi et al. [15]. Therefore, assynthesized MCM-48 was calcined at 823 K in air for 5 h to remove the occluded surfactant. Then, 1 g (16 mmol SiO2 ) of calcined and dehydrated MCM-48 was stirred vigorously in 25 ml
of dry toluene containing 1.6 mmol of APTES (TEOS/APTES
molar ratio 1/10) at 383 K for 10 h. The functionalized material was collected by filtration, washed several times with isopropanol, and dried at 373 K. The material was then acidified
using an HCl solution in order to protonate the amino groups
leading to positively charged ammonium moieties. Therefore,

100 mg of amino-functionalized MCM-48 was stirred in 100 ml

of 0.1 M HCl for 6 h at ambient temperature as described by
Yokoi et al. [15]. The acidified material, designated thereafter as
MCM-48NH+
3 , was recovered by filtration and dried at room
temperature overnight before its use as an adsorbent.
2.3. Materials characterization
Nitrogen adsorptiondesorption experiments were performed at 77 K using a Quantachrome Autosorb 1 volumetric analyzer. Prior to analysis, the materials were degassed under
vacuum until a residual pressure of 105 mbar was reached
at 200 C for 4 h for the pure silica materials and 120 C
for 12 h for the functionalized materials. The specific surface area, SBET , was determined from the linear part of the
BET plot (P /P0 = 0.050.20). The pore size distributions
were calculated from the desorption branch using the Barrett
JoynerHallenda and the HorvathKawazoe methods for the
mesoporous (pore size 2.0 nm) and microporous (pore size
<2.0 nm) materials, respectively [16,17]. The total pore volume
was evaluated from the adsorbed amount at a relative pressure
of about 0.99.
X-ray powder diffraction of the synthesized materials was
carried out using a Rigaku D-Max-Ultima III diffractometer
with nickel-filtered CuK radiation of wavelength 1.5406 .
Powder diffraction patterns were obtained between 0.75 and
10 with a scan speed of 1 /min.
Fourier transform infrared (FTIR) spectra were collected between 200 and 4000 cm1 on a Varian 1000 (Scimitar series)
spectrometer with 2 cm1 resolution using KBr pellets at room
temperature.
2.4. Adsorption studies
The adsorption capability of the MCM-48NH+
3 toward
nitrate and phosphate anions was investigated separately using aqueous solutions of NaNO3 and NaH2 PO4 . Adsorption
was performed batch-wise in Erlenmeyer flasks mounted on a
temperature-regulated platform stirrer under the following conditions: temperature 545 C, adsorbent loading 110 g/L, anion initial concentration 0.10.7 g/L, and pH 210. The pH of
the solutions was adjusted by addition of 0.01 M solution of
HCl or NaOH. For all the adsorption tests presented hereafter,
the stirring was continued for 60 min before aliquots of the solutions were withdrawn, filtered with a microsyringe (0.2 m),
and analyzed for nitrate or phosphate anion using an ion chromatograph ICS 2500 from Dionex equipped with an electrochemical detector and an IonPac AS18 (4 250 mm) column.
The mobile phase was a 35 mM KOH solution.
The percentage removal (R) of anions was calculated as
R=

Ci Ce
100%.
Ci

(1)

The amount of anions retained on the sorbent phase (mg/g) was

calculated as
Qe = (Ci Ce )

V
,
m

(2)

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

where Ci and Ce are the initial and final concentrations of the

anion in solution (mg/L), V is the solution volume (L), and m
is the mass of sorbent (g).
The adsorption experiments were performed in duplicate or
triplicate and analyzed twice.
2.5. Ion adsorption competitivity
The competitive aspect was studied for the adsorption of
either nitrate or phosphate anions from separate aqueous so
lutions of 100 mg NO
3 or H2 PO4 /L in the presence of equivalent concentrations of the following competitive species: Cl ,

2+
2+
SO2
4 , HCO3 , Ca , and Mg .
2.6. Desorption and regeneration studies
Desorption tests were conducted in NaOH solutions. Due
to the possible dissolution of silica in NaOH solutions, the effects of NaOH concentration on the dissolution of silica was
first investigated. To this purpose, calcined MCM-48 samples
were stirred in NaOH solutions at concentrations ranging from
0.01 to 0.1 M for 1 h. The undissolved silica materials were
recovered by filtration, dried, and weighed. The dissolution of
silica in NaOH solution occurs according to the following reaction [18]:
SiO2 + 2NaOH Na2 SiO3 + H2 O.

377

3. Results and discussion

3.1. Adsorbent characterization
The pure silica MCM-48 exhibited a typical reversible nitrogen adsorptiondesorption isotherm of type IV, having an
S-shape characteristic of mesoporous materials. The MCM-48
NH2 showed an isotherm with an adsorption step being shifted
toward lower relative pressures (Fig. 1a). As presented in Table 1, the BET surface area decreased from 1400 m2 /g for
pure silica material to 890 m2 /g for the MCM-48NH2 . The
corresponding pore sizes were 2.2 and 1.9 nm (see Fig. 1b)
and pore volumes were 1.04 and 0.66 cm3 /g for the MCM-48
and MCM-48NH2 , respectively. The decrease in the textural
properties was previously reported to accompany surface modification of mesoporous materials with organic functional groups
[11] and was attributed to the occupation of the pore framework by the functional groups protruding into the internal pore
surface. Thus, to prevent complete pore blockage and drastic
decrease in the surface area, which may affect the adsorbent
performances, the organosilane/pure silica ratio was voluntarily fixed to 10% during the synthesis of MCM-48NH2 .

(3)

The desorption kinetics investigations were conducted on spent

adsorbent samples stemming from tests performed with initial
anion concentrations of 300 mg/L and adsorbent loading of
5 g/L, leading to an average amount adsorbed (Qe(i) ) of 32
and 34 mg/g for nitrate and phosphate, respectively. The spent
adsorbents were agitated in 0.01 M NaOH solutions at 5 g/L
adsorbent loading. Aliquots were sampled at different time intervals, filtered, and analyzed by ionic HPLC. The percentage
of desorbed (D) nitrate or phosphate anions was calculated as


Qe(des)
D= 1
(4)
100%,
Qe(i)
where Qe(i) represents the initial amount adsorbed (mg/g) on
the sample subjected to desorption and Qe(des) represents the
amount desorbed (mg/g) during the desorption in NaOH solution.
Consecutive adsorptiondesorption cycles were also performed to study the regenerability and stability of MCM-48
NH+
3 . Hence, after contacting a fresh adsorbent with nitrate
or phosphate solution (300 mg/L), it was separated by filtration, washed, dried at room temperature, and weighed. Then,
the anion-loaded sample was stirred in 0.01 M NaOH solution
for 1 h with a loading of 5 g adsorbent/L to strip the nitrate or
phosphate anions. After filtration and washing with deionized
water, the solid was dried at room temperature and weighed.
It was then acidified in HCl solution as described above in the
synthesis of MCM-48NH+
3 section. These steps comprise the
first adsorptiondesorption cycle, which was followed by four
more cycles using the same adsorbent batch.

(a)

(b)
Fig. 1. Nitrogen adsorptiondesorption isotherms of MCM-48 materials.

378

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

Table 1
Textural properties of calcined and functionalized MCM-48
Adsorbent

D211 spacing
(nm)

BET surface area

(m2 /g)

Pore diameter
(nm)

Total pore volume

(cm3 /g)

Calcined MCM-48
MCM-48NH2
MCM-48NH+
3

3.25
3.24
3.10

1400
890
750

2.2
1.9
1.2

1.04
0.66
0.39

Fig. 3. Infrared spectra for MCM-48 materials.

Fig. 2. X-ray diffraction patterns for MCM-48 materials.

Furthermore, nitrogen adsorption performed on the MCM48NH+

3 revealed a subsequent decrease in its textural properties and a change in the isotherm shape from type IV to type I
characteristic of microporous materials (Figs. 1a and 1b and Table 1). This may be attributed to the progressive pore filling of
the material by the counter-anion Cl , which balances the positive charge formed during the protonation using HCl solution.
Powder X-ray diffraction patterns are depicted in Fig. 2.
MCM-48, MCM-48NH2 , and MCM-48NH+
3 all exhibited
typical XRD patterns of cubic I a3d mesophase with characteristic Bragg peaks around 2.8 and 3.0 attributed to the
(211) (220) diffraction lines, respectively. This indicates that
the modified materials retained the ordered cubic structure even
though the MCM-48NH+
3 diffractogram denoted a slight shift
toward higher values for the two main peaks, thus confirming
the shrinking in the pore diameters, in agreement with the nitrogen adsorption data.
FTIR spectra of MCM-48, MCM-48NH2 , and MCM-48
NH+
3 are shown in Fig. 3. In both spectra, strong absorption bands near 12001000 cm1 were observed and attributed
to SiOSi asymmetric stretching vibrations. The broad band
within the 34003600 cm1 interval encompassing several
not-well-resolved vibrations was ascribed to hydrogen-bonded
SiOH groups [19,20]. For the pure silica MCM-48, the vibration observed at 3747 cm1 is assigned to free SiOH groups
[20,21]. A huge attenuation of this vibration in the amine and
ammonium functionalized spectra indicates that most of the
free silanol groups participated in the functionalization reaction. Vibrations at 2945, 2877, and 1469 cm1 observed for
the MCM-48NH2 spectrum were ascribed to CH asymmetric

stretching, CH symmetric stretching, and CH2 scissor vibrations, respectively [21]. Fading of the intensity of these vibrations in MCM-48NH+
3 suggests that acidification could affect
the absorption of CH and CH2 groups. Vibration at 1650 cm1
in both spectra is attributed to H2 O molecules [20]. Furthermore, vibration around 1600 cm1 observed only in MCM-48
+
NH2 and MCM-48NH+
3 spectra is assigned to NH2 or NH3
groups [20].
3.2. Adsorption tests
Preliminary adsorption tests performed within 2 h with sampling intervals of 2 min showed that the adsorption took place
rapidly and the equilibrium was reached within 10 min of contact time for both nitrate and phosphate anions. Therefore, for
the tests presented thereafter, the stirring was continued for
60 min before aliquots were withdrawn and analyzed.
3.3. Effects of adsorbent loading and anion initial
concentration
The percentage removal of nitrate and phosphate from their
respective solutions at different initial concentrations increased
with the increase in the adsorbent loading (see Table 2). This
effect is due to the increase in the total available adsorbent surface area and adsorption sites with increasing adsorbent loading. Furthermore, an increase in initial anion concentrations
decreased the percentage removal of nitrate and phosphate for
the same adsorbent loading. Moreover, the percentages removal
at equilibrium, reached in the presence of phosphate anions,
were higher than those obtained with the nitrate anions, indicat-

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

379

Table 2
Effects of adsorbent loading and initial concentration on nitrate and phosphate adsorptive removals (%) at 25 C (adsorption time: 1 h)
Adsorbent loading
(g/L)

100

Nitrate initial concentration (mg/L)

200

300

500

700

2
5
7
10

38.1 0.1
60.6 1.8
64.5 1.7
71.7 0.5

29.3 5.2
51.5 3.5
52.7 4.8
67.0 4.6

16.4 5.2
43.0 2.4
52.2 0.8
60.7 1.6

9.2 4.9
24.4 2.8
37.9 0.1
50.5 5.1

7.2 4.3
18.8 0.1
24.2 1.9
28.2 2.5

Phosphate initial concentration (mg/L)

2
5
7
10

100

200

300

500

700

50.0 3.2
81.0 0.7
88.0 2.2
90.0 1.6

38.0 1.6
66.0 0.4
74.0 1.6
88.0 4.7

23.0 4.1
55.0 5.5
61.0 4.4
70.0 5.2

14.0 2.8
37.0 4.4
50.0 5.1
58.0 0.1

8.0 2.0
24.0 1.0
36.0 0.1
45.0 0.3

ing that the ammonium sites present on the acidified adsorbents

have a better affinity to the phosphate anions. This effect can be
explained by the two different anion structures. Indeed, the negative charge on the nitrate anion is delocalized between three
oxygen atoms [18]. This delocalization could suggest that the
charge on each oxygen atom is 1/3. In contrast, the charge
1 is localized on one oxygen atom of the phosphate anion.
Because each oxygen atom of the nitrate ion has a smaller individual charge, it has less ability to stabilize this charge [22], explaining the lower migration to positively charged ammonium
sites and hence the smaller percentage of nitrate adsorbed. It
should be noted that for both nitrate and phosphate adsorption,
the amounts adsorbed were not proportional to the adsorbent
loadings. For instance, with an initial nitrate concentration of
200 mg/L, adsorbent loadings of 2 and 10 g/L led to removals
of 29.3 and 67.0%, respectively.
3.4. Effect of pH
The pH of solution is an important parameter in the investigated adsorption process. Indeed, pH influences the charge
of the grafted propylamine groups on the mesoporous silica
walls, the pKa of which is 10.6 [23]. Hence, when pH <
pKa , the amino groups are charged positively upon protonation, leading to ammonium moieties capable of attracting anionic species by electrostatic forces. Conversely, when pH >
pKa , the alkyl-ammonium species are deprotonated and recover
their free amine form, unable to exert any electrostatic effect
toward anionic species. On the other hand, phosphate anions oc2
3
cur as monovalent H2 PO
4 , divalent HPO4 , or trivalent PO4
forms depending on the solution pH (pK1 = 2.2, pK2 = 7.2,
pK3 = 12.4) [24].
Fig. 4 shows the variation in the amount adsorbed as a function of initial pH for nitrate and phosphate anions. In the case
of nitrate, the percentage removal remained constant up to pH
8, at which the amino groups are still positively charged. Beyond this limit, the adsorption performances decreased significantly. As for phosphate, the uptake denoted three zones. It
increased up to a maximum of 34 mg/g, reached at pH 5, then
it decreased drastically, reaching a plateau around 12 mg/g at
pH  8. This could be attributed to the surface charge of the adsorbent becoming neutral at higher pH and losing electrostatic

Fig. 4. Effect of pH on nitrate and phosphate anion adsorption on MCM-48

NH+
3 at 25 C (anion initial concentration 300 mg/L, adsorbent loading 5 g/L,
adsorption time 1 h). Lines show trends. Bars show standard deviations of measurements.

attraction toward the occurring negatively charged phosphate

3
anions HPO2
4 and PO4 [25]. Thus, the optimal pH for phosphate removal ranged between 4.0 and 6.0, at which the dominant phosphate form is the monovalent H2 PO
4.
3.5. Adsorption isotherms
The adsorption isotherm studies were conducted to probe the
effects of temperature on the adsorption of nitrate and phosphate anions on MCM-48NH+
3 and to determine the maximum
adsorption capacities. Hence, the adsorption isotherms for nitrate and phosphate anions on MCM-48NH+
3 carried out at
different temperatures are depicted in Figs. 5 and 6, respectively. Application of the Langmuir model for both adsorption
data allowed the determination of the maximum adsorption capacities, which were, for instance, ca. 43 and 46 mg/g at the
lowest temperature investigated (5 C) for nitrate and phosphate anions, respectively. Accordingly, the adsorbed amounts
at equilibrium decreased with increasing temperature, indicating the exothermic character of the adsorption process. Furthermore, MCM-48NH+
3 outperformed other adsorbents tested

380

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

Table 3
Effects of competitive ions on the adsorption of nitrate and phosphate anions
Coexisting ionsa

Amount adsorbedb (mg/g)

NO
3 alone

NO
3 + Cl
2
NO
+
SO
3
4

12.2 0.3
6.9 1.3
6.7 1.0

NO
3 + HCO3

Fig. 5. Experimental adsorption isotherms on MCM-48NH+

3 for nitrate with
5 g/L adsorbent loading. Lines show trends.

2+
NO
3 + Ca
2+
NO
+
Mg
3

H2 PO4 alone

H2 PO
4 + Cl
2
H2 PO
+
SO
4
4

H2 PO4 + HCO
3
2+
H2 PO
4 + Ca

2+
H2 PO4 + Mg

10.3 0.4
7.5 0.6
10.0 1.4
16.5 0.6
14.9 0.1
9.5 0.7
17.3 0.1
15.1 0.1
12.7 0.7

a Operating conditions: nitrate or phosphate initial concentration 100 mg/L,

competitive ion initial concentration 100 mg/L, adsorbent loading 5 g/L, temperature 25 C, adsorption time 60 min.
b The amount adsorbed refers accordingly to nitrate or phosphate.

Fig. 6. Experimental adsorption isotherms on MCM-48NH+

3 for phosphate
with 5 g/L adsorbent loading. Lines show trends.

under comparable conditions such as quaternized biomass anion exchanger, which exhibited maximum adsorption capac
ities of 10 mg NO
3 /g and 16 mg H2 PO4 /g [26]. On the
other hand, poly(allylamine hydrochloride) polymer hydrogels
demonstrated adsorption capacities of 15 mg NO
3 /g and 17 mg
H2 PO
/g
[27].
4
3.6. Effect of competitive ions
The effect of the presence of competitive anions or cations
are presented in Table 3. In the case of nitrate, the adsorbed
amount decreased by ca. 45% in the presence of chloride and
sulfate, while in the presence of magnesium and carbonate, a
decrease of about 20% of the amount adsorbed was observed.
As for phosphate, the presence of chloride, calcium, and magnesium ions did not disturb the adsorption significantly, while
the presence of sulfate decreased the adsorbed amount by about
43% if compared to the monoionic solution. However, adsorption of phosphate increased slightly in the presence of carbonate
ions. This can be attributed to the solution pH equilibrating at
4.0 in the presence of carbonate anions, which value was previously shown to be optimal for phosphate adsorption.

Fig. 7. Dissolution of adsorbent in sodium hydroxide solution. Adsorbent loading: 5 g/L. Lines show trends.

3.7. Desorption studies and regeneration

The dissolution profile of MCM-48 mesoporous silica in
NaOH solutions is depicted in Fig. 7. As seen, silica was almost completely dissolved for NaOH concentrations exceeding 0.04 M. While the dissolution was almost negligible below a NaOH concentration of 0.01, it monotonically increased
with increasing NaOH concentration within the interval 0.01
0.04 M. Therefore, sodium hydroxide at 0.01 M was considered
as an upper limit to prevent the adsorbent dissolution and was
employed for desorption studies. Fig. 8 shows the desorption
kinetics relative to spent adsorbents used for nitrate and phosphate adsorption. As seen, desorption was rapid and complete
for both anions within 10 min. This leads to the conclusion that
the adsorption of anions on MCM-48NH2 is reversible. On
the other hand, consecutive adsorptiondesorption cycle data
are shown in Table 4. As seen, no significant loss of adsorbent capacity was observed during five cycles of adsorption
desorption, demonstrating that the adsorbent was very suitable
for the design of a continuous sorption process.

R. Saad et al. / Journal of Colloid and Interface Science 311 (2007) 375381

381

Table 4
Regeneration of the adsorbent
Cycle
1
2
3
4
5

a
NO
4

a
H2 PO
4

% Removal

Amount adsorbed (mg/g)

% Removal

Amount adsorbed (mg/g)

46.0 3.7
41.5 8.7
40.5 4.0
40.0 9.4
43.3 4.7

27.6 2.2
24.9 5.2
24.3 2.4
24.0 5.6
26.0 2.8

59.0 6.1
60.6 3.3
52.5 3.5
57.5 5.4
56.6 1.8

35.4 3.7
36.4 1.9
31.5 2.1
34.5 3.2
33.4 1.0

a Operating conditions: nitrate or phosphate initial concentration 300 mg/L, adsorbent loading 5 g/L, temperature 25 C, adsorption time 60 min.

for Innovation is gratefully acknowledged. The authors thanks

Professor L.E. Parent for the use of the Varian IR spectrometer.
References

Fig. 8. Desorption kinetics of nitrate and phosphate anions at 25 C in 0.01 M

NaOH solution. Adsorbent loading: 5 g/L. Lines show trends.

4. Conclusion
The results reported in this study allow us to conclude that
ammonium-functionalized MCM-48 represents a promising adsorbent for the removal of nitrate and monovalent phosphate
anions from aqueous solutions even at high concentrations

(700 mg/L NO
3 or H2 PO4 ). Adsorption capacity increased
with the adsorbent loading. For both nitrate and phosphate anions, the adsorption was shown to be negatively affected by
increasing temperature. The removal of nitrate was optimal at
pH < 8, while for phosphate, the adsorption was optimal in the
pH interval of 46. The desorption was rapid and it was completely achieved within 10 min for both anions using NaOH
at 0.01 M. Regeneration tests showed that no significant loss
of adsorbent capacity was observed during five adsorption
desorption cycles.
Acknowledgments
Funding provided by the Natural Sciences and Engineering Research Council of Canada and the Canada Foundation

[1] S.E. Jorgensen, Water quality: The impact of eutrophication, UN Environment Programme, 2001.
[2] W. Viessman Jr., M.J. Hammer, Water Supply and Pollution Control, seventh ed., Pearson Prentice Hall, Upper Saddle River, NJ, 2005.
[3] S.N. Levine, D.W. Schindler, Can. J. Fish. Aquat. Sci. 46 (1989) 2.
[4] World Health Organization, Health hazards from nitrate in drinking water,
Environ Health, vol. 1, WHO, Copenhagen, 1985.
[5] A.P. Murphy, Nature 350 (1991) 223.
[6] N.M. Agyei, C.A. Strydom, J.H. Potgieter, Cem. Concr. Res. 32 (2002)
1889.
[7] Z.H. Li, R.S. Bowman, Water Res. 35 (2001) 3771.
[8] M. Ozacar, Chemosphere 51 (2003) 321.
[9] D.Y. Zhao, A.K. Sengupta, Water Res. 32 (1998) 1613.
[10] U.S. Orlando, A.U. Baes, W. Nishijima, M. Okada, Chemosphere 48
(2002) 1041.
[11] A. Sayari, S. Hamoudi, Chem. Mater. 13 (2001) 3151.
[12] A. Stein, B.J. Melde, R.C. Schroden, Adv. Mater. 12 (2000) 1403.
[13] H. Yoshitake, T. Yokoi, T. Tatsumi, Chem. Mater. 14 (2002) 4603.
[14] K. Schumacher, M. Grn, K. Unger, Micropor. Mesopor. Mater. 27 (1999)
201.
[15] T. Yokoi, H. Yoshitake, T. Tatsumi, J. Mater. Chem. 14 (2004) 951.
[16] G. Horvath, K. Kawazoe, J. Chem. Eng. Jpn. 16 (1983) 470.
[17] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373.
[18] G.I. Brown, Introduction to Inorganic Chemistry, second ed., Longman,
Harlow/New York, 1985, pp. 215244.
[19] H. Yoshitake, T. Yokoi, T. Tatsumi, Bull. Chem. Soc. Jpn. 76 (2003) 847.
[20] L. Bois, A. Bonhomm, A. Ribes, B. Pais, G. Raffin, F. Tessier, Colloids
Surf. A Physicochem. Eng. Aspects 221 (2003) 221.
[21] X. Zhao, G. Lu, A. Whittaker, G. Millar, J. Phys. Chem. 101 (1997) 6525.
[22] T.A. Geissman, Principles of Organic Chemistry, Freeman, San Francisco,
1959, pp. 170190.
[23] D.C. Harris, Quantitative Chemical Analysis, second ed., Freeman, New
York, 1987, pp. 729738.
[24] D.D. Perrin, B. Dempsey, Buffers for pH and Metal Ion Control, Chapman
& Hall, London, 1974, pp. 156162.
[25] S. Karaca, A. Grses, M. Ejder, M. Aikyildiz, J. Hazard. Mater. 128
(2006) 273.
[26] C. Namasivayam, W.H. Holl, J. Chem. Technol. Biotechnol. 80 (2005)
164.
[27] D.R. Kioussis, F.W. Wheaton, P. Kofinas, Aquat. Eng. 23 (2000) 315.